Thoracic irradiation (RT) is a major component of therapy for many patients with tumors of the breast, lung, esophagus, thymus and mediastinal lymphatics. In these patients, RT-induced pulmonary injury is one of the most common treatment-related toxicities. Despite the large number of patients that receive thoracic RT, the biologic and physical determinants of RT-induced lung injury are not fully understood. There are no accepted means of predicting, prior to delivery of RT, what the physiologic consequences of the RT will be. In the prior funding period, we successfully exploited functional lung imaging and three-dimensional (3D) radiation treatment planning tools to address this issue. Using single photon emission computed tomography (SPECT) lung perfusion scans as an assessment of regional function, we defined the dose-response curve for RT-induced regional lung injury, and its temporal nature. Preliminary data suggests that RT- induced changes in whole lung function can be quantitatively related to the sum of regional injuries. We established the importance of understanding the RT dose distribution relative to the functioning (i.e., perfused) regions of the lung in addition to the anatomical CT-defined lung, and the importance of pre-RT whole lung function, in predicting RT-induced whole lung injury. Based on our work, we have developed algorithms that should be predictive for RT-induced changes in lung function. Furthermore, we and others have demonstrated that TGF-beta (transforming growth factor-beta) may be a marker to identify patients who are predisposed to develop RT-induced lung injury. The current proposal builds on our prior work to address persistent/new issues. First, we will prospectively test the ability of our algorithms (based on the dose distribution within the anatomic and functional lung, and pre-RT pulmonary function) to predict which patients are at relatively high vs. low risk of experiencing RT-induced reductions in whole lung function. Second, we will further study the relationship between regional lung injury (assessed with SPECT perfusion scans) and changes in whole lung function. Whole lung function will be assessed, as done previously, by changes in pulmonary function tests and pulmonary symptoms. In addition, we will assess exercise tolerance by monitoring changes m the distance the patient can walk in six minutes. This six minute walk distance is a more clinically and functionally relevant metric since it measures overall exercise capacity. Third, we will prospectively assess the impact of TGF-beta on RT- induced lung dysfunction. This project will increase understanding of the dosimetric, physiologic and biologic determinants of RT-induced lung injury. The development, testing and validation of such predictive algorithms will improve patient care by reducing the incidence of lung injury and facilitate "safe" dose escalation for intrathoracic cancers.